2021, 68 (6), 691-699

Original

Regulation of 26 member 7 (Slc26a7) by thyroid stimulating hormone in thyrocytes

Yuta Tanimura1), 2), Mitsuo Kiriya1), Akira Kawashima1), Hitomi Mori1), Yuqian Luo1), 3), Tetsuo Kondo2) and Koichi Suzuki1)

1) Department of Clinical Laboratory Science, Faculty of Medical Technology, Teikyo University, Itabashi, Tokyo 173-8605, Japan 2) Department of Pathology, Faculty of Medicine, University of Yamanashi, Chuo, Yamanashi 409-3898, Japan 3) Department of Laboratory Medicine, Nanjing Drum Tower Hospital and Jiangsu Key Laboratory for Molecular Medicine, Nanjing University Medical School, Nanjing 210008, China

Abstract. Iodine transportation is an important step in thyroid hormone biosynthesis. Uptake of iodine into the thyroid follicle is mediated mainly by the basolateral sodium–iodide symporter (NIS or solute carrier family 5 member 5: SLC5A5), and iodine efflux across the apical membrane into the follicular lumen is mediated by (SLC26A4). In addition to these transporters, SLC26A7, which has recently been identified as a causative for congenital hypothyroidism, was found to encode a novel apical iodine transporter in the thyroid. Although SLC5A5 and SLC26A4 have been well-characterized, little is known about SLC26A7, including its regulation by TSH, the central hormone regulator of thyroid function. Using rat thyroid FRTL-5 cells, we showed that the mRNA levels of Slc26a7 and Slc26a4, two apical iodine transporters responsible for iodine efflux, were suppressed by TSH, whereas the mRNA level of Slc5a5 was induced. Forskolin and dibutyryl cAMP (dbcAMP) had the same effect as that of TSH on the mRNA levels of these transporters. TSH, forskolin and dbcAMP also had suppressive effects on SLC26A7 promoter activity, as assessed by luciferase reporter gene assays, and levels, as determined by Western blot analysis. TSH, forskolin and dbcAMP also induced strong localization of Slc26a7 to the cell membrane according to immunofluorescence staining and confocal laser scanning microscopy. Together, these results suggest that TSH suppresses the expression level of Slc26a7 but induces its accumulation at the cell membrane, where it functions as an iodine transporter.

Key words: SLC26A7, SLC5A5, SLC26A4, TSH, Thyroid

IODINE is a fundamental element for the biosynthesis circulating iodine into thyrocytes. SLC26A4 (or pendrin), of the thyroid hormones T3 and T4, and the prime func‐ another iodine transporter, was identified as a causative tion of the thyroid is to concentrate iodine and make it gene for in 1997, and the protein available for the biosynthesis of thyroid hormones. localizes on the apical membrane, opposite to SLC5A5 Iodine derived from dietary sources is transported into [2-4]. SLC26A7, encoding another member of the same the thyrocyte across the basal membrane and then transporter family as SLC26A4, was demonstrated to be a secreted across the apical membrane into the follicular novel causative gene for congenital hypothyroidism in lumen where it is organified to the tyrosine residues of 2018 [5-7]. Although SLC26A7 was first identified as a thyroglobulin, a large protein stored in the follicular chloride–bicarbonate anion exchanger in the kidney and colloid. Thus, the thyroid follicle serves as the minimal stomach [8], expression profiling of different human functional unit of hormone synthesis and secretion. tissues showed predominant mRNA expression of Sodium–iodide symporter (NIS or solute carrier SLC26A7 in the thyroid. Immunofluorescence staining of family 5 member 5: SLC5A5), the first iodine transporter thyroid tissue sections showed that SLC26A7 is local‐ cloned, in 1996 [1], is localized on the basolateral ized at the apical membrane, similar to pendrin [3, 5, 9]. membrane of thyrocytes and is responsible for taking up Bicameral culture system and radioiodine transport/ uptake studies suggested that SLC26A7 functions in the Submitted Aug. 8, 2020; Accepted Jan. 15, 2021 as EJ20-0502 efflux of iodine in culture medium [5], similar to pendrin Released online in J-STAGE as advance publication Feb. 14, 2021 [10]. SLC26A7 was thus considered a novel apical iodine Correspondence to: Koichi Suzuki, Ph.D., Department of Clinical Laboratory Science, Faculty of Medical Technology, Teikyo transporter responsible for iodine efflux in the thyroid. University, 2-11-1 Kaga, Itabashi, Tokyo 173-8605, Japan. TSH is the primary regulator of iodine uptake and E-mail: [email protected] stimulator of thyroid hormone production and secretion.

©The Japan Endocrine Society 692 Tanimura et al.

TSH activates the Slc5a5 promoter to induce the expres‐ Slc26a7 reverse, 5'-CAAGCCACCTGTGTCTTTGC-3'; sion and membrane localization of the protein [9, 11-14]. NM_017008.4 Gapdh forward, 5'-ACAGCAACAGGGT It also induced the membrane abundance of SLC26A4 GGTGGAC-3'; Gapdh reverse, 5'-TTTGAGGGTGCAG and enhanced its function of iodine efflux into the folli‐ CGAACTT-3'. cular lumen, without affecting the total protein level [15, 16]. Although, the effects of TSH on SLC5A5 and Protein preparation and Western blot analysis SLC26A4 have been evaluated extensively, the role of Preparation of cellular and Western blot TSH in the regulation of the newly identified SLC26A7 analysis were performed as described previously [19, 22, is not known. Here, we investigated the potential effect 25]. Briefly, cells were washed with ice-cold PBS and of TSH on Slc26a7 with respect to its expression and lysed in lysis buffer containing 20 mM Tris-HCl (pH subcellular localization in rat thyroid FRTL-5 cells. 7.5), 150 mM NaCl, 1 mM Na2EDTA, 1 mM EGTA, 1% NP-40, 1% sodium deoxycholate, 2.5 mM sodium pyro‐

Materials and Methods phosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4 and 1 μg/mL leupeptin. Cells were then scraped using a Cell culture and treatment disposable cell lifter (Corning, Corning, NY, USA) and Rat thyroid FRTL-5 cells were grown in Coon’s centrifugated. The supernatant was recovered, and pro‐ modified Ham’s F-12 medium supplemented with 5% tein concentrations were measured using the DC Protein bovine serum (Invitrogen, Waltham, MA, USA) and Assay Kit (Bio-Rad Laboratories, Hercules, CA, USA). a mixture of six hormones (1 mU/mL bovine TSH, Ten micrograms of each protein sample were separated 10 μg/mL insulin, 0.36 ng/mL hydrocortisone, 5 μg/mL by electrophoresis on NuPage 4–12% Bis-Tris gels transferrin, 10 ng/mL somatostatin and 2 ng/mL glycyl- (Invitrogen) and transferred to a nitrocellulose membrane L-histidyl-L-lysine acetate) as described previously using i-Blot gel transfer stacks (Invitrogen). The mem‐ [17-19]. After the cells attached to the bottom of plates brane was washed with 0.1% Tween 20 in PBS (PBST), (Greiner Bio-One, Kremsmünster, Austria), they were placed in blocking buffer (PBST containing 5% nonfat maintained in the same medium without TSH for 5 days, milk) for 1 h, and then incubated with mouse anti- followed by exposure to TSH, forskolin or dibutyryl Slc26a7 (1:1,000; Novus Biologicals, Centennial, CO, cAMP (dbcAMP; Sigma-Aldrich, St. Louis, MO, USA). USA) or mouse anti-β-actin (1:5,000; Cell Signaling Technology, Danvers, MA, USA) overnight. After Total RNA isolation and real-time PCR washing with PBST, the membrane was incubated with Total RNA was purified using the RNeasy Plus Mini horseradish peroxidase (HRP)-labelled horse anti-mouse Kit (Qiagen, Hilden, Germany), and cDNA was synthe‐ IgG (1:1,000; Cell Signaling Technology) for 1 h. The sized using the High-Capacity cDNA Reverse Transcrip‐ HRP was detected by chemiluminescence using the tion Kit (Applied Biosystems, Waltham, MA, USA) as ImmunoStar LD reagents (FUJIFILM Wako Pure Chem‐ described previously [20-22]. Real-time PCR was per‐ ical Corporation, Osaka, Japan) and analyzed by the C- formed using Fast SYBR Green Master Mix (Applied DiGit blot scanner (both from LI-COR, Lincoln, NE, Biosystems) according to the manufacturer’s instructions USA). and run on the Thermal Cycler Dice Real Time System III (Takara Bio, Tokyo, Japan). Briefly, 5 ng cDNA was Luciferase reporter gene assay mixed with 10 μL 2× Fast SYBR Green Master Mix and Among four transcript variants of SLC26A7, we used amplified by incubating for 30 sec at 95°C, followed by one that has the most upstream transcription start site 40 cycles of 5 sec at 95°C and 60 sec at 60°C, and 1 (NM_001282356.2). A series of sequences within the 5'- cycle of 15 sec at 95°C, 30 sec at 60°C and 15 sec at flanking region of the human SLC26A7 gene, starting at 95°C for dissociation. Real-time PCR was conducted in positions –2,207, –1,653, –1,140 and –443, relative to at least triplicate, and the relative mRNA expression the transcription start site, were amplified by PCR from levels were normalized to the corresponding Gapdh level human genomic DNA using the following forward and using the ∆∆Ct method as described [19, 23, 24]. The reverse primer pairs containing NheI or XhoI restriction sequences of the PCR primers are as follows: enzyme sites: –2,207 forward, 5'-CGTGCTAGCTGTGG NM_052983.2 Slc5a5 forward, 5'-CTACCGTGGGTGG CTAGGGATC-3'; –1,653 forward, 5'- CGTGCTAGCTA TATGAAGG-3'; Slc5a5 reverse, 5'-TGCCACCCACTAT GGGCTGTATAACAAAGTAT-3'; –1,140 forward, 5'- GAAAGTCC-3'; NM_019214.1 Slc26a4 forward, 5'- CGTGCTAGCATCTACTCAGTAGGGTGAG-3'; –443 CAAGTGGGTTCTTGCCTCCT-3'; Slc26a4 reverse, 5'- forward, 5'-CGTGCTAGCAGCACTGAAAAATCA TTGGTGGCGTAGACTTTCCC-3'; XM_008763531.2 A-3'; and reverse, 5'-GATCTCGAGACTTCCGTATTCT Slc26a7 forward, 5'-TGCTCCCCCAATGAACATCC-3'; TACCAGG-3'. The PCR products were introduced into Regulation of Slc26a7 in the thyroid 693 the pGL3-Basic luciferase reporter plasmid (Promega, PCR results clearly showed that TSH significantly sup‐ Madison, WI, USA) via the NheI and XhoI sites. FRTL-5 pressed Slc26a7 mRNA levels in a concentration- cells (2 × 105) maintained without TSH for 5 days in 6- dependent manner (Fig. 1A, left panel); on the other well plates (Greiner Bio-One) were transfected with luci‐ hand, TSH significantly induced Slc5a5 mRNA expres‐ ferase reporter plasmids using FuGENE HD transfection sion (Fig. 1A, middle panel), which has been demons‐ reagent (Promega). At 6 h after transfection, TSH, trated previously [12, 26]. The Slc26a4 mRNA level forskolin or dbcAMP was added to the culture medium, was also reduced by TSH (Fig. 1A, right panel), similar and cells were maintained for 48 h. Cells were harvested, to the change in Slc26a7 mRNA. Such changes in and a luciferase assay was performed using the Bright- mRNA levels were observed as early as 6 h after TSH Glo Luciferase Assay System (Promega). Luciferase stimulation and progressed until 24 h (Fig. 1B). These activity was measured using the ARVO MX plate reader results suggest that TSH induces mRNA expression of (PerkinElmer, Waltham, MA, USA) and normalized to Slc5a5, the basolateral iodine transporter responsible for the corresponding protein concentrations determined iodine uptake, while TSH suppresses expression of the using the DC Protein Assay Kit (Bio-Rad Laboratories). apical iodine transporters, Slc26a7 and Slc26a4, respon‐ sible for iodine efflux. Immunofluorescence staining and confocal laser scanning microscopy Forskolin and dbcAMP reproduced the suppressive After attaching to the bottom of 4-well chamber slides effect of TSH on Slc26a7 mRNA expression (Thermo Fisher Scientific, Waltham, MA, USA), To further evaluate the observed action of TSH, we FRTL-5 cells were maintained in the absence of TSH for examined the effects of forskolin (which increases the 5 days. Cells were then treated with 1 mU/mL TSH for intracellular cAMP concentration by stimulating adenyl‐ 6, 12, 24, 48 and 120 h, and 1 μM forskolin or 500 μM ate cyclase [27]) and dbcAMP (a cell-permeable cAMP dbcAMP for 120 h. After incubation, cells were fixed analog) on the Slc26a7 mRNA level. Real-time PCR using 10% neutral buffered formalin for 10 min at room revealed that both forskolin and dbcAMP suppressed the temperature, and then permeabilized with PBS contain‐ mRNA level of Slc26a7 in a concentration-dependent ing 0.1% Triton X-100. After blocking with PBS con‐ manner at 24 h (Fig. 2, left panels), which was similar to taining 3% bovine serum albumin, the slides were the effect of TSH, shown in Fig. 1. Forskolin and incubated with mouse anti- Slc26a7 (1:500; Novus dbcAMP induced Slc5a5 mRNA expression (Fig. 2, Biologicals) overnight, washed with PBST and then middle panels) and reduced Slc26a4 mRNA expression incubated with Alexa Fluor 594-labeled goat anti-mouse (Fig. 2, right panels), also reproducing the effects of IgG (1:1,000; Cell Signaling Technology), followed by TSH. These results confirmed that the TSH/cAMP sig‐ nuclear staining with Hoechst 33342 (Cell Signaling naling cascade increases the expression of Slc5a5, Technology). Fluorescence was observed under a confo‐ encoding a basolateral iodine transporter, while the same cal laser scanning microscope (FV10i-DOC, Olympus, signal suppressed the expression of Slc26a7 and Slc26a4, Tokyo, Japan). apical iodine transporters, in thyrocytes.

Statistical analysis TSH suppressed Slc26a7 protein expression in All experiments were performed in triplicate and FRTL-5 cells repeated at least three times. Results are expressed as the We next examined the effect of TSH on Slc26a7 pro‐ mean ± standard deviation (SD). Significant differences tein expression using Western blot analysis. FRTL-5 were determined by one-way ANOVA followed by cells cultivated in the absence of TSH for 5 days were Dunnet’s post hoc test. p < 0.05 was considered to repre‐ exposed to TSH, forskolin or dbcAMP at increasing con‐ sent significance. centrations for 24 h. Western blot analysis of the total cellular protein content clearly revealed that TSH, for‐ Results skolin and dbcAMP all suppressed the Slc26a7 protein level in a concentration-dependent manner (Fig. 3, upper TSH suppressed Slc26a7 mRNA expression in panels). Densitometric analysis of the specific bands FRTL-5 cells confirmed the significant reductions in the Slc26a7 pro‐ We first examined the potential effect of TSH on the tein level relative to the β-actin level by TSH, forskolin Slc26a7 mRNA level using real-time PCR. After cultur‐ and dbcAMP (Fig. 3, lower panels). ing FRTL-5 cells in medium without TSH for 5 days, fresh medium supplemented with 0.01, 0.1 or 1 mU/mL TSH suppressed human SLC26A7 promoter activity TSH was added to the cells for up to 24 h. The real-time To investigate whether TSH/cAMP signaling directly 694 Tanimura et al.

Fig. 1 TSH suppresses Slc26a7 mRNA expression in concentration- and time-dependent manners in FRTL-5 cells FRTL-5 cells were maintained in basal medium without TSH for 5 days and then switched to medium containing increasing concentrations of TSH (0.01, 0.1 and 1 mU/mL) for 24 h (A). FRTL-5 cells maintained in basal medium without TSH for 5 days were switched to medium containing 1 mU/mL TSH for up to 24 h (B). Total RNA was extracted and subjected to real-time PCR to determine the relative mRNA levels of Slc26a7, Slc5a5, Slc26a4 and Gapdh. Relative mRNA levels were normalized to the Gapdh level and expressed relative to the control level (n = 3). *p < 0.05; ** p < 0.01; *** p < 0.001, compared with the control. affects the transcription of SLC26A7, we cloned 5'- any of the conditions tested (Fig. 4). flanking regions of human SLC26A7 and performed luci‐ These results led to speculation that the promoter ferase reporter gene assays to evaluate the promoter sequence between positions –2,207 and –1,653 is neces‐ activity. FRTL-5 cells maintained in the absence of TSH sary for TSH regulation of SLC26A7. This, in silico for 5 days were transfected with SLC26A7 promoter/luci‐ analysis was performed to predict potential transcription ferase chimeric plasmids or the control plasmid for 6 h. binding sites using PROMO (http://alggen.lsi.upc.es/cgi- Then, the culture medium was replaced with fresh bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3) and medium supplemented with TSH, forskolin or dbcAMP, CiiiDER (http://ciiider.com/). The results revealed that and the cells were cultured for 48 h before luciferase three NK2 homeobox 1 (NKX2-1; also known as thyroid assays were performed. Among the four SLC26A7 pro‐ transcription factor-1: TTF-1) binding sites are located in moter constructs, that harboring the sequence extending the 5'-flanking region of SLC26A7, from positions to position –2,207 of the 5'-flanking region showed –2,207 to –1,653 (–2,090 to –2,085, –2,077 to –2,072, strong basal promoter activity in the absence of TSH, –2,051 to –2,046). In addition, the 5'-flanking region which was almost comparable with the activity of the contained apparent TATA and CpG box as biological pGL3 Control plasmid driven by the strong SV40 signals of core promoter elements. Therefore, NKX2-1 promoter and enhancer (Fig. 4, control). Whereas the might be responsible for mediating the effects of TSH/ promoter activity of the pGL3 Control plasmid was cAMP signaling on SLC26A7 promoter activation, which enhanced by TSH and not significantly affected by is similar to the transcriptional regulation of other forskolin or dbcAMP under the conditions tested, the thyroid-specific such as Tg, TPO, SLC5A5 and activity of the SLC26A7 promoter was significantly sup‐ TSHR [14, 28, 29]. pressed by TSH, forskolin and dbcAMP. These results suggest that the observed decrease in the Slc26a7 mRNA TSH induced membrane localization of Slc26a7 in level was a result of suppression of SLC26A7 promoter FRTL-5 cells activity. No significant transcriptional activity was As an apical iodine transporter, cell membrane locali‐ detected in the other luciferase reporter constructs under zation is expected to be crucial for the function of Regulation of Slc26a7 in the thyroid 695

Fig. 2 Both forskolin and dbcAMP suppress Slc26a7 mRNA expression in a concentration-dependent manner in FRTL-5 cells FRTL-5 cells were maintained in basal medium without TSH for 5 days and then switched to medium containing increasing concentrations of forskolin (0.01, 0.1 and 1 μM) or dbcAMP (50, 200 and 500 μM) for 24 h. Total RNA was extracted and subjected to real-time PCR to determine the relative mRNA levels of Slc26a7, Slc5a5, Slc26a4 and Gapdh. Relative mRNA levels were normalized to the Gapdh level and expressed relative to the control level (n = 3). * p < 0.05; ** p < 0.01; *** p < 0.001, compared with the control.

Fig. 3 TSH, forskolin and dbcAMP suppress Slc26a7 protein expression in a concentration-dependent manner in FRTL-5 cells FRTL-5 cells were maintained in basal medium without TSH for 5 days and then switched to medium containing increasing concentrations of TSH (0.01, 0.1 and 1 mU/mL), forskolin (0.01, 0.1 and 1 μM) or dbcAMP (50, 200 and 500 μM) for 24 h. Total cellular proteins were extracted and subjected to Western blot analysis to determine the protein levels of Slc26a7 and β-actin (upper panels). The densitometric values of the specific bands are presented as bar graphs and represent the relative protein levels normalized to the β-actin level (lower panels).

Slc26a7 [5]. To demonstrate the potential effects of TSH stimulation resulted in Slc26a7 moving from the perinu‐ on the subcellular localization of Slc26a7, we analyzed clear area to the cytoplasm (Fig. 5, 6–12 h) and then Slc26a7 localization in FRTL-5 cells by immunofluores‐ gradually moving to the cell membrane (Fig. 5, 24–120 cence staining and confocal laser scanning microscopy. h). The fluorescence intensity of Slc26a7 in cells was In basal medium without TSH, Slc26a7 protein was reduced by TSH with time, which is in agreement with observed in the cytoplasm in a somewhat granular pat‐ the decrease in the total Slc26a7 protein level tern especially around the nucleus (Fig. 5, 0 h). TSH demonstrated by Western blot analysis, shown in Fig. 3. 696 Tanimura et al.

Fig. 4 TSH, forskolin and dbcAMP suppress human SLC26A7 promoter activity (A) FRTL-5 cells maintained in basal medium without TSH were transfected with the pGL3-Basic luciferase reporter plasmid carrying 5'-flanking regions of human SLC26A7 of the indicated lengths, or with the pGL3 Control plasmid. The cells were then treated with 1 mU/mL TSH, 1 μM forskolin or 500 μM dbcAMP for 48 h. Cells were harvested, and luciferase assays were performed. Promoter activity is expressed as mean ± SD relative light units (n = 3). All values were normalized to the total amount of cellular protein. (B) Schematic representation of SLC26A7 promoter region. Nkx2-1 binding site, GC box, CAAT box and TATA box are shown relative to the transcription start site (TSS).

Furthermore, both forskolin and dbcAMP exerted the suppressive effect of TSH on the SLC26A7 promoter similar effect as TSH to translocate Slc26a7 protein on may be, at least in part, mediated by down-regulation of the cell membrane (Fig. 5). Together, these results sug‐ NKX2-1 expression via TSH. Detailed promoter analysis gest that TSH/cAMP signaling induces translocation of is needed to elucidate the whole picture of transcriptional the Slc26a7 protein to the plasma membrane, where it regulation of SLC26A7 gene. functions as a transporter, while suppressing its total While suppressing the total cellular protein level of cellular expression level. Slc26a7, TSH/cAMP exerted a completely different effect to concentrate Slc26a7 on the plasma membrane. Discussion When cells were maintained in TSH-free culture medium, the Slc26a7 protein was localized mainly The present study aimed to elucidate the effects of around nuclei. After stimulation with TSH, Slc26a7 TSH on the expression and subcellular localization of gradually translocated into the cytoplasm and then to the Slc26a7, a newly identified apical iodine transporter, in plasma membrane, as demonstrated by confocal laser thyrocytes. We showed that stimulation of FRTL-5 cells scanning microscopy. Interestingly, this is similar to the with TSH, forskolin or dbcAMP suppressed Slc26a7 observed changes in Slc5a5 localization induced by mRNA expression, as well as the activity of the TSH, in which Slc5a5 localized to intracellular compart‐ SLC26A7 promoter, which contains three putative ments throughout the cytoplasm in the absence of TSH NKX2-1 biding sites. It was previously shown that the but localized to the plasma membrane in the presence of transcription of Slc26a4 was positively regulated by TSH [12]. Regarding SLC26A4, we previously showed Nkx2-1 [30], and that TSH decreased Nkx2-1 protein that SLC26A4 proteins harboring mutations (L236P, expression as well as its association with the promoters T416P and G384E), identified in patients with Pendred of its target genes in FRTL-5 cells [31]. Therefore, the syndrome exhibiting defective iodine organification, Regulation of Slc26a7 in the thyroid 697 were retained in the endoplasmic reticulum and not seen at the plasma membrane [32]. Similarly, in vitro studies showed that over-expressed FLAG-conjugated wildtype SLC26A7 was detected mainly at the cell membrane, where it exported intracellular iodine outside of the cell, whereas a mutant SLC26A7 protein aggregated in the cytosol instead of localizing to the cell membrane and no longer exhibited iodine transporter function [5]. Taken together, the migration of SLC26A7 from the cytoplasm to the plasma membrane induced by TSH suggests acti‐ vation of its transporter function. Future electrophysio‐ logical studies are needed to determine whether TSH actually enhances the transporter function of SLC26A7, along with its membrane localization. We showed a similarity between the regulation of Slc26a7 and Slc26a4, two apical iodine transporters, by TSH. Previous studies showed that TSH suppressed the gene expression and induced membrane localization of Slc26a4 [15, 16], as observed with Slc26a7 in this study. In rat thyroid PCCL-3 cells, TSH and forskolin rapidly increased Slc26a4 abundance at the plasma membrane via the TSH/cAMP/PKA signaling cascade [16], which was correlated with an increase in iodine efflux [16]. It is not clear why the expression levels of the apical iodine transporters Slc26a7 and Slc26a4 are decreased, but their membrane localization is induced, by TSH. Since FRTL-5 cells are monolayer cells without having polari‐ zation, studies of the effect of iodine on Slc26a7 expres‐ sion using polarized cells may provide a clearer picture for understanding the fine-tuning of iodine-transporting mechanisms. Belonging to the same transporter family, Slc26a7 and Slc26a4 share a common protein structure. Two-thirds of the NH2 terminus is largely hydrophobic and encom‐ passes the transmembrane segments, in which the anion- binding sites and pores are putatively located. The COOH- terminal domain of the protein is predicted to contain a and anti-sigma factor antagonist (STAS) domain [33]. Studies on Arabidopsis thaliana SLC26 transporters and disease-causing mutations in the STAS domain of SLC26A3 suggest that this domain is important for the membrane anchoring and transporter Fig. 5 TSH induces translocation of Slc26a7 to the plasma function of SLC26 family members [34]. Elimination membrane in FRTL-5 cells of the distal portion of the STAS domain containing a FRTL-5 cells were maintained in basal medium without putative PKA phosphorylation site (RKDT714–717) in TSH for 5 days and then switched to medium containing 1 mU/mL TSH, 1 μM forskolin or 500 μM dbcAMP for SLC26A4 resulted in partial loss of function and the indicated times. Cells were fixed and subjected to decreased membrane localization in response to TSH or immunofluorescent staining of Slc26a7 (red) and nuclear forskolin [15, 16]. Further, in Japanese patients with (blue) staining using Hoechst 33342. Cells were then goitrous congenital hypothyroidism, a p.Gln500Ter examined under a confocal laser scanning microscope. mutation was identified in the STAS domain of Bar = 10 μm. SLC26A7 [5], and expression of this mutant protein in HEK293T cells caused abnormal localization in the cyto‐ sol and loss of iodine transporter function [5]. 698 Tanimura et al.

Taken together, our findings and those from other gesting that the transporter function is activated by TSH. groups suggest that the TSH signaling cascade may These results encourage further studies to explore the induce Slc26a7 membrane localization via phosphoryla‐ underlying mechanisms by which TSH regulates the sub‐ tion of the STAS domain of Slc26a7. This post- cellular localization along with the transporter function transcriptional modification, in addition to transcriptional of Slc26a7. regulation, is an important way for TSH to regulate the function of SLC5A5 [12], SLC26A4 [15, 16] and most Acknowledgements likely SLC26A7 as well. It also suggests that the mem‐ brane localization and transporter function of SLC26A7 The authors would like to thank Ms. Madoka Naito, might be pronounced in Graves’ thyroid, in which TSHR Teikyo University for her secretarial assistance. This is constantly stimulated by the thyroid-stimulating anti‐ work was supported by JSPS KAKENHI Grant Number body (TSAb), similar to the strong membrane staining of JP15K09444, JP17K08990 and JP19K07875. SLC26A4 in thyroid tissue sections from patients with Graves’ disease [3, 9]. Disclosure In summary, we demonstrated that the TSH/cAMP signaling cascade suppressed the promoter activity and None of the authors have any potential conflicts of expression level of Slc26a7 in FRTL-5 cells. However, interest associated with this research. TSH induced membrane localization of Slc26a7, sug‐

References

1. Dai G, Levy O, Carrasco N (1996) Cloning and character‐ 9. Kondo T, Nakamura N, Suzuki K, Murata S, Muramatsu ization of the thyroid iodide transporter. Nature 379: 458– A, et al. (2003) Expression of human pendrin in diseased 460. thyroids. J Histochem Cytochem 51: 167–173. 2. Everett LA, Glaser B, Beck JC, Idol JR, Buchs A, et al. 10. Yoshida A, Taniguchi S, Hisatome I, Royaux IE, Green (1997) Pendred syndrome is caused by mutations in a ED, et al. (2002) Pendrin is an iodide-specific apical por‐ putative sulphate transporter gene (PDS). Nat Genet 17: ter responsible for iodide efflux from thyroid cells. J Clin 411–422. Endocrinol Metab 87: 3356–3361. 3. Royaux IE, Suzuki K, Mori A, Katoh R, Everett LA, et al. 11. Kogai T, Endo T, Saito T, Miyazaki A, Kawaguchi A, et (2000) Pendrin, the protein encoded by the Pendred syn‐ al. (1997) Regulation by thyroid-stimulating hormone of drome gene (PDS), is an apical porter of iodide in the thy‐ sodium/iodide symporter gene expression and protein lev‐ roid and is regulated by thyroglobulin in FRTL-5 cells. els in FRTL-5 cells. Endocrinology 138: 2227–2232. Endocrinology 141: 839–845. 12. Riedel C, Levy O, Carrasco N (2001) Post-transcriptional 4. Bidart JM, Mian C, Lazar V, Russo D, Filetti S, et al. regulation of the sodium/iodide symporter by thyrotropin. (2000) Expression of pendrin and the Pendred syndrome J Biol Chem 276: 21458–21463. (PDS) gene in human thyroid tissues. J Clin Endocrinol 13. Kogai T, Taki K, Brent GA (2006) Enhancement of Metab 85: 2028–2033. sodium/iodide symporter expression in thyroid and breast 5. Ishii J, Suzuki A, Kimura T, Tateyama M, Tanaka T, et al. cancer. Endocr Relat Cancer 13: 797–826. (2019) Congenital goitrous hypothyroidism is caused by 14. Endo T, Kaneshige M, Nakazato M, Ohmori M, Harii N, dysfunction of the iodide transporter SLC26A7. Commun et al. (1997) Thyroid transcription factor-1 activates the Biol 2: 270. promoter activity of rat thyroid Na+/I– symporter gene. 6. Cangul H, Liao XH, Schoenmakers E, Kero J, Barone S, Mol Endocrinol 11: 1747–1755. et al. (2018) Homozygous loss-of-function mutations in 15. Bizhanova A, Chew TL, Khuon S, Kopp P (2011) Analy‐ SLC26A7 cause goitrous congenital hypothyroidism. JCI sis of cellular localization and function of carboxy- Insight 3: e99631. terminal mutants of pendrin. Cell Physiol Biochem 28: 7. Zou M, Alzahrani AS, Al-Odaib A, Alqahtani MA, 423–434. Babiker O, et al. (2018) Molecular analysis of congenital 16. Pesce L, Bizhanova A, Caraballo JC, Westphal W, Butti hypothyroidism in Saudi Arabia: SLC26A7 mutation is a ML, et al. (2012) TSH regulates pendrin membrane abun‐ novel defect in thyroid dyshormonogenesis. J Clin Endo‐ dance and enhances iodide efflux in thyroid cells. Endo‐ crinol Metab 103: 1889–1898. crinology 153: 512–521. 8. Petrovic S, Barone S, Xu J, Conforti L, Ma L, et al. (2004) 17. Suzuki K, Lavaroni S, Mori A, Ohta M, Saito J, et al. – – SLC26A7: a basolateral Cl /HCO3 exchanger specific to (1998) Autoregulation of thyroid-specific gene transcrip‐ intercalated cells of the outer medullary collecting duct. tion by thyroglobulin. Proc Natl Acad Sci U S A 95: 8251– Am J Physiol Renal Physiol 286: F161–F169. 8256. Regulation of Slc26a7 in the thyroid 699

18. Oda K, Luo Y, Yoshihara A, Ishido Y, Sekihata K, et al. messenger RNA and protein expression. J Endocrinol (2017) Follicular thyroglobulin induces cathepsin H 167: 125–135. expression and activity in thyrocytes. Biochem Biophys 27. Fradkin JE, Cook GH, Kilhoffer MC, Wolff J (1982) For‐ Res Commun 483: 541–546. skolin stimulation of thyroid adenylate cyclase and cyclic 19. Yoshihara A, Hara T, Kawashima A, Akama T, Tanigawa 3',5'-adenosine monophosphate accumulation. Endocrinol‐ K, et al. (2012) Regulation of dual oxidase expression and ogy 111: 849–856.

H2O2 production by thyroglobulin. Thyroid 22: 1054– 28. Civitareale D, Castelli MP, Falasca P, Saiardi A (1993) 1062. Thyroid transcription factor 1 activates the promoter of the 20. Suzuki K, Mori A, Saito J, Moriyama E, Ullianich L, et al. thyrotropin receptor gene. Mol Endocrinol 7: 1589–1595. (1999) Follicular thyroglobulin suppresses iodide uptake 29. Francis-Lang H, Price M, Polycarpou-Schwarz M, Di by suppressing expression of the sodium/iodide symporter Lauro R (1992) Cell-type-specific expression of the rat gene. Endocrinology 140: 5422–5430. thyroperoxidase promoter indicates common mechanisms 21. Yoshihara A, Luo Y, Ishido Y, Usukura K, Oda K, et al. for thyroid-specific gene expression. Mol Cell Biol 12: (2019) Inhibitory effects of methimazole and propyl‐ 576–588. thiouracil on iodotyrosine deiodinase 1 in thyrocytes. 30. Dentice M, Luongo C, Elefante A, Ambrosio R, Salzano Endocr J 66: 349–357. S, et al. (2005) Pendrin is a novel in vivo downstream tar‐ 22. Ishido Y, Yamazaki K, Kammori M, Sugishita Y, Luo Y, get gene of the TTF-1/Nkx-2.1 homeodomain transcrip‐ et al. (2014) Thyroglobulin suppresses thyroid-specific tion factor in differentiated thyroid cells. Mol Cell Biol 25: gene expression in cultures of normal but not neoplastic 10171–10182. human thyroid follicular cells. J Clin Endocrinol Metab 31. Medina DL, Suzuki K, Pietrarelli M, Okajima F, Kohn 99: E694–E702. LD, et al. (2000) Role of insulin and serum on thyrotropin 23. Akama T, Sue M, Kawashima A, Wu H, Tanigawa K, et regulation of thyroid transcription factor-1 and pax-8 al. (2012) Identification of microRNAs that mediate thy‐ genes expression in FRTL-5 thyroid cells. Thyroid 10: roid cell growth induced by TSH. Mol Endocrinol 26: 295–303. 493–501. 32. Rotman-Pikielny P, Hirschberg K, Maruvada P, Suzuki K, 24. Sellitti DF, Puggina E, Lagranha C, Doi S, Pithon-Curi T, Royaux IE, et al. (2002) Retention of pendrin in the endo‐ et al. (2007) TGF-beta-like transcriptional effects of thy‐ plasmic reticulum is a major mechanism for Pendred syn‐ roglobulin (Tg) in mouse mesangial cells. Endocr J 54: drome. Hum Mol Genet 11: 2625–2633. 449–458. 33. Aravind L, Koonin EV (2000) The STAS domain—a link 25. Suzuki K, Kohn LD (2006) Differential regulation of api‐ between anion transporters and antisigma-factor antago‐ cal and basal iodide transporters in the thyroid by thyro‐ nists. Curr Biol 10: R53–R55. globulin. J Endocrinol 189: 247–255. 34. Shibagaki N, Grossman AR (2006) The role of the STAS 26. Kogai T, Curcio F, Hyman S, Cornford EM, Brent GA, et domain in the function and biogenesis of a sulfate trans‐ al. (2000) Induction of follicle formation in long-term cul‐ porter as probed by random mutagenesis. J Biol Chem tured normal human thyroid cells treated with thyrotropin 281: 22964–22973. stimulates iodide uptake but not sodium/iodide symporter